On-line fault diagnosis of hydraulic systems using Unscented Kalman Filter

This paper presents a scheme that automatically generates hydraulic system fault symptoms by on-line processing of raw sensor data from a real hydraulic test rig. A condition-monitoring scheme based on the Unscented Kalman Filter (UKF) is developed on a validated model. The UKF algorithm estimates the system states and generates the residual errors. Four types of faults, which cannot be directly detected from current sensor values, are investigated in this work: external chamber leakage at either side of the actuator, internal leakage between the two hydraulic cylinder chambers, dynamic friction load, sudden loss of load and increment in load. Also, for each leakage scenario, three levels of leakage are used in the experiments. The developed UKF-based fault monitoring scheme is tested on the practical system while different fault scenarios are individually introduced to the system.